[Hadoop] HDFS Reading

The Reading operation in HDFS?

• In HDFS the main two operations are Reading and Writing data using HDFS.
• To read from HDFS, at first, the client sends requests for metadata to the NameNode. When the NameNode responds, then it sends a response with the number of blocks, their location, different information about replicas and some other details.
• Now the client communicates with the DataNodes. The client can read data in the parallel fashion. When the whole data is read, it combines the blocks as the original file.

• This is the pictorial representation of the Data Reading procedure.